In medical image registration, for example, images of a certain modality have to be transformed in the coordinate system of an image of another modality which might be taken at another point in time. Registration is necessary in order to be able to compare or integrate the images obtained from different modalities.
When a three-dimensional multi-modality dataset derived for example by magnetic resonance (MR) or computed tomography (CT), which is representing an anatomical region is going to be blended with live two dimensional x-ray projection images (e.g. for the purpose of 3D roadmapping), it is necessary that the mapping (transformation model) between the frame of registration of the x-ray system (e.g. a C-arm x-ray system) and the frame of registration of the three-dimensional multi-modality dataset is known.
The process of obtaining this mapping is called “registration”. The direct registration of the three-dimensional dataset to the two-dimensional x-ray projection image can be performed either manually, or automatically, employing a registration algorithm.
With respect to the results achieved with the present registration systems there is still the desire to improve the results achieved in the registration process.